186 Proceedings of the Royal Society of Edinburgh. [Sess. 
experienced here again from the values got from the commercial metal. 
Two lines, with totally different values of (a) were required, the values 
being : 
A. a = 15 B. a = 55 
n='i'2 n = *33 
6 = -681 b = 104. 
In the commercial specimen one line was found sufficient, with a = 65, 
n — *45, 6 = T66. 
Aluminium Wire — Effect of Antecedent Heating or Cooling. 
A series of experiments was carried out with wires of aluminium 
raised to increasing temperatures and cooled in the same manner as has 
been described in the former paper in the case of brass. The wires were 
heated in an electric furnace, and the range taken was between the 
ordinary temperature and 560° C., the temperatures being recorded as 
before by a platinum thermometer. In each case two values of a were 
